you are viewing a single comment's thread
view the rest of the comments
[–] 3 points 1 week ago (2 children)

Ugh, was bound to happen. We need open models, BADLY. Unfortunately one also needs the hardware for them. I only have 20 GB VRAM and 32 GB of RAM, basically can't use a decent model with that, sucks.

  • source
  • hideshow 2 child comments
  • [–] 2 points 1 week ago

    You can just spin up the compute for pretty cheap. It's not as nice as having a local model, for sure, but until the bubble pops and all this hardware floods the market, or we can start grabbing it after they're done with it, it'll have to do.

  • source
  • parent
  • [–] 1 point 1 week ago

    Have you tried qwen 3.8 flash next? So far I've tested it only briefly but it's definitely a lot better compared to other local llms that I can run (~30B class models). For reference I am on 16 vram and 32 ram, older ddr4 platform.

  • source
  • parent